We propose to selectively assay mutations in two separate but related systems using structure-guided approaches. One system is based on the structure of the cyclin-CDK (cyclin dependent kinase) complex, whereas the other system is nucleic-acid based. Both systems concern critical regulatory events in the cell cycle of the budding yeast S. cerevisiae, which has proven to be an excellent model for the regulation of the cell cycle in all eukaryotes. Both approaches heavily rely on using MidasPlus to visualize known structures. Careful analysis of these structures will suggest mutations to be made and assayed in vivo. First, we will study the structure of the cyclin-CDK complex (PDB entry 1FIN) as solved by Jeffery, Paveltich and colleagues. In particular, we will look for residues on the surface on the cyclin that are near the ATP substrate in the kinase. These residues may be involved in substrate specificity. Further, by creating mutations in yeast cyclins at homologous positions, we will look for cyclins which still activate the CDK CDC28 for H1 kinase activity but fail to activate CDC28 in vivo because they cannot target CDC28 specifically to its physiological substrates. We will also attempt to make dominant negative cyclins which still bind CDC28 but which fail o activate it because they do not reposition the T loop, or again, because they fail to activate CDC28 specifically for a subset of substrates. Creating these mutants will involve studying the known crystal structures of cyclin-CDK and CDK alone with MidasPlus to identify critical residues and to appreciate the conformational changes. We speculate that dominant negative cyclins can be designed as potent CDK inhibitors that could be delivered by gene therapy. Second, we are studying the promoter of the HO endonuclease of Saccharomyces cerevisiae because it is highly activated at the G1-S transition of the cell cycle by the transcription factors SWI4 and SWI6. Furthermore, a small region of the HO promoter, URS2, is able to confer cell cycle specific regulation on heterologous promoters. In particular, fully silenced by URS2 until the G1-S transition. Preliminary mutagenesis studies indicate that the overall topology and bend in this AT rich promoter sequence creates a unique structure that is inactive in the absence of SWI4 and SWI6 proteins. SWI4 and SWI6 may convert this structure to an active state by DNA bending. We propose to use MidasPlus to study known structures of DNA, both intrinsically bent DNA and DNA-transcription factor cocrystals in which the DNA is bent upon factor binding. These studies should suggest mutations or permutations that could be created in URS2 to disrupt the structure and thus the function of this region. Given that we have a simple biological assay for URS2 function in yeast, we can readily test the hypotheses formulated by studying DNA structures using MidasPlus. RR01 081-20